Homogeneous nucleation of vapor from polymer solutions

John H. Jennings, and Stanley Middleman. Macromolecules , 1985, 18 (11), pp 2274–2276. DOI: 10.1021/ma00153a037. Publication Date: November 1985...
0 downloads 0 Views 341KB Size
Macromolecules 1985, 18, 2274-2276

2274

Homogeneous Nucleation of Vapor from Polymer Solutions John H. Jenningst and Stanley Middleman*$ Chemistry Department and Chemical Engineering-AMES Department, The University of California, Sun Diego, La Jolla, California 92093. Received January 22,1985 ABSTRACT The limit of superheat is measured for solutions of polystyrene in cyclohexane at atmospheric pressure. The presence of polymer leads to increases in superheat of the order of 10 "C, to an extent that depends upon both molecular weight and concentration.

Introduction If a liquid is carefully isolated from surfaces that might provide sources of heterogeneous nucleation of vapor, i.e., normal boiling, then it is possible to raise the temperature of the liquid well above its normal boiling point without the occurrence of a vapor phase. For pure liquids, classical nucleation theory1 predicts that the limit of superheat is about 88-91% of the critical temperature, and experimental data for a variety of liquids confirm this.2 We are interested in the effect on the limit of superheat of adding soluble polymer to a pure liquid. As we shall demonstrate, the addition of high molecular weight polymer increases the limit of superheat to a degree that varies with molecular weight and concentration. Classical nucleation theory is presented and discussed in a number of sources. For example, Blander and K a t 9 present an expression of the nucleation rate J (number of bubbles formed per cm3 per second) in the form

J = 3.73 X lo3"(

W B

ex.[

-1.182 x 1052 T(Pv - PL)

Symbols are defined in the Nomenclature section, but we point out here that the numerical coefficients given in eq 1require use of the units presented with these symbols. In writing eq 1 we have adopted the common approximation that, a t atmospheric pressure, the Poynting correction is unity. As others have pointed out, the exponential term dominates the variation of J with parameters, and over a very narrow temperature range J increases from a small number (corresponding to unlikely observation of nucleation) to a very large number. Because the temperature range is narrow one defines a single nucleation temperature, or limiting superheat T,, and observations of nucleation (as mentioned above) consistently suggest that T,= 0.89Tc. If physical property data are available one may calculate the value of J a t T, and find values in the range 104-106 bubbles cm-3 s-l. Further, the value of J changes by about 3 orders of magnitude per degree Celsius in the neighborhood of T,. As Reid has pointed out,l0 no kinetic theory is available to treat nonideal liquid mixtures, of which polymer solutions are certainly an example. A stability theory may also be developed with thermodynamic arguments and has been successfully applied and tested for pure liquids and ideal mixtures. No such extension to polymer solutions has been carried out. Hence there is no theoretical basis upon which our results can be compared or discussed.

Measurements The limit of superheat is measured by the standard rising-drop meth~dA . ~small drop of the test liquid is introduced through a syringe at the bottom of a column fiied with a denser immiscible Chemistry Department.

* Chemical Engineering-AMES

Department.

0024-9291/85/2218-2214$01.50/0

liquid. The drop rises through the column under the influence of buoyancy. The column is transparent glass and is wrapped helically along its length with resistance wire. By controlling the pitch of the helix and the current through the wire, one may impose a temperature profile, increasing upward, in the host liquid in the column. By choosing a relatively viscous host liquid, one may suppress convection and thermal mixing due to drop motion. When proper care is taken one may observe a drop that rises, and in doing so heats up, until the drop reaches the limiting superheat T,. At this point the drop explodes and creates a vapor bubble. One repeats such experiments until a thermocouple can be positioned in the host liquid at the level where nucleation is consistently observed. Experiments were performed first with pure organic solvents and their mixtures, in order to check the data against literature values. Then solutions of polystyrene in cyclohexane were prepared. In all cases, glycerol served as the host liquid. Polysciences polystyrene was used. Four monodisperse samples were studied, at molecular weights of 2 X lo3, 4 X lo3, 50 X lo3, and 100 x IO3. Dilute solutions were prepared with Mallinckrodt AR grade cyclohexane. These solutions were then filtered through a 0.2-pm Fluoropore filter made by Millipore Corp. From this stock, concentrated solutions were prepared by evaporating solvent by passing filtered air over the dilute solutions. Solutions were introduced through a septum at the bottom of the column with a Hamilton microsyringe. Drop volumes of 0.1, 0.2, and 0.4 p L were studied. The limit of superheat was found to be independent of drop size in this range, an indication that no appreciable diffusion of solvent into the host phase was occurring.

Results Figure 1 shows data obtained with mixtures of pentane and cyclohexane. The observed limiting superheat is plotted against mass fraction. Mole fraction is more commonly used in presenting data for liquid mixtures. However, since molecular weight of some polymers may be uncertain, or since many polymers are not available in monodisperse samples, we choose to present these results using mass fraction. Our data differ slightly from those of Holden and Katz4 for the pure solvents; we are 2 "C low for pure pentane and 3 "C high for pure cyclohexane. On the other hand, our data are within 1 "C of the values presented by Blander.* Figure 2 shows data for low molecular weight polystyrenes in cyclohexane. Also shown is a line connecting the values of limiting superheat for cyclohexane and pure styrene, which we may regard as the lowest molecular weight polystyrene. A single data point obtained for a 5.5% styrene/cyclohexane solution is in good agreement with this curve. Out to 30% polymer the variation of T, with polymer concentration is linear when plotted in this form. Figure 3 shows data obtained with high molecular weight polymers. At a molecular weight of 50000 the data are similar to those for lower molecular weight samples, although it does not appear that the dependence is linear a t low concentrations. The highest molecular weight studied, 100 000, shows the most interesting behavior. It appears that the limiting 1985 American Chemical Society

Macromolecules, Vol. 18, No. 11, 1985

Nucleation of Vapor from Polymer Solutions 2275

-

\

503

97,200

D

+-

493

1

400,000